INDEX

Explanations

references to specific geographical locations, particularly cities and capitals

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 11-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L11R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.11.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

atte

-0.20

xr

-0.15

tura

-0.14

edReader

-0.14

adera

-0.14

orna

-0.14

ave

-0.14

face

-0.14

erner

-0.14

POSITIVE LOGITS

redients

0.16

asmus

0.16

byn

0.15

 personals

0.14

eration

0.13

 legis

0.13

ãĤīãģĽ

0.13

PIO

0.13

 civil

0.13

_mB

0.13

Activations Density 0.088%

references to specific geographical locations, particularly cities and capitals

No Comments

No Known Activations